Some Definitions: Wiki

Wiki is hot right now because of Wikipedia’s 1 million article achievement. This has lead to some confusion between wiki (wikiwiki=quick, quick), a collaborative software concept, and the Wikipedia, the most visible implementation of that concept. For the record: wiki is free editing of pages and keeping revision histories. It is merely a way to rapidly compose a set of non-hierarchical, associated web pages (like an encyclopedia). Other criteria will fall down immediately and are for zealots: allowing edits without signing in (Wikipedia might end the practice); open membership ( Jotspot is a cool corporate application); a specific look, feel or affiliation (the main reason I run MediaWiki is that I can’t abide WikiNames).

This is a bit disingenuous, though, since those two simple features, free edit and revision history, are all-important. Wiki works well because it is an unsophisticated content management system. There is only one rule regarding pages: in general, if it’s a page, you can edit it. In the past, the rule has been:if it’s a page, and the owner gives you permission, or if you own it, you can edit it. Also, because wiki relies so heavily on a “social hack”, it’s hard to imagine wiki without The Wiki Way, the set of infocommunistic mores that hold Wikipedia and other wikis together. Wikipedia solves the previously mentioned design challenges in an elegant and, from most people’s natural command-and-control impulses, non-intuitive way: by letting people assign themselves to tasks, disavowing ownership, and allowing disagreements to resolve themselves through discussion and edit wars.

The success of Wikipedia speaks to the virtues of open systems, but following the same theme as the last entry, I wonder whether open systems don’t favor volunteerism and the ability to donate large amounts of (leisure) time. The Wikipedia has grown as big as it has because early on it established the precedent of a large number of people donating small portions of spare time. The sight of a large volunteer effort provides the initial inducement to contribute. But the underlying reality is that a smaller but substantial group of “marginal” users donate large amounts of time and are responsible for a disproportionate bulk of the contributions.

These people have their reasons, not the least of which is the pleasure they get from writing and receiving positive feedback, but the disparities of output between marginal and normal users would be a problem in a wiki less grounded in volunteerism. Imagine a company wiki where several employees are responsible for maintaining important documentation. A marginal user in the workplace would want to make his larger number of edits known, and would resent sharing the credit for a well-written page with someone whose contribution consisted of cursory spell-checking. This might be considered a problem with capitalism itself, but a similar problem could arise with an academic wiki written by people who consider their time extremely valuable.

As I’ve put more and more time into my own labor-of-love wiki (why, I’m the #1 contributor!), I’ve felt greater misgivings about the prospect of other people adding to or in any way sharing ownership of “my” work. I know that this is not the Wiki Way, but I can’t help feeling somewhat proprietary about my words. Examining my own fragile ego, I think that a sufficient social hack would be to add the “Created by:” label some wikis have, along with an automatically generated credits page that would list percentage of overall and retained contributions made by a user to a page.

I’ve also had a desire for a version-freeze system, so that a page could be frozen at 1.0, 2.0 etc., with the frozen versions being the public face of a document while revisions go on behind the scenes (similar to the CVS development cycle). There are other things that can be done with wiki, including the introduction of complex group decision-making and evaluation processes (e.g., voting and mod points), although these would detract from the current simplicity and openness. And in general, version history is simply a good idea for documents, especially if splitting is tracked in a consistent way. For example, I wrote this blog entry as if it were a wiki, which is why it’s so disorganized.